Revel M16 Speaker Review

[amirm]

The dips in the vertical response around the crossover point have nothing to do with floor and ceiling reflections. These are indicators of the lobe behavior - a wrong observation that has been made in several of the other speaker reviews. The woofer is sourced from China, not Indonesia (SB Acoustics).

Come again??? No one has said the floor or the ceiling are the cause. What is said is the to the lobing that exists, you want to reduce those reflections. Once there, their effect is lowered.

 

[amirm]

Understood.
A wild idea: maybe you can make a list of top5 most promising speakers that you think are a really good investment of your time and money and allow members to chip in with cash for those that they'd like to see bought & tested. Given that you need only a single speaker and even a single tower can have a good resell value, it would probably take just 10-20 members with $50 each to cover the majority of the cost for a good mid-tier speaker. I'm definitely in for 2-3 speakers now and more later. As a side benefit you'll know what your audience really wants to see the most. Voting with money is the best voting. Cheers.

I have no such list to offer. What would be fun is to test some very high end speakers (Wilson, Evolution Acoustics, YG, etc.). None of you can help with purchases of this magnitude. We will get to test them at some point in the future once we have a strong voice in this market through owners loaning them to me.

As for funding, it makes no sense for me to have my hand out for this effort and that effort. I spend money on running the site, shipping things back and forth, buying gear to test, and paying off the capital cost of the expensive instrumentation. Indeed this is why I did not try to fund the speaker instrumentation by getting money for it.

You all can get together however, make your lists, fund them together, buy the speaker for me to test. Once there, you can figure out how to sell and get your money back. I have hardly sold anything I have bought and am terrible at doing so.

As for my audience, 3/4th of them are not even members so can't be represented with any poll. We have nearly one million page views a month! Polls are useful but ultimately I have no hope that you can satisfy even a small portion of membership with any vote for this or that speaker.

Finally, I have tons of speakers here for testing. As I am typing this, another is being measured. And more is coming. So sitting here, supply of speakers is not a problem for me.

 

[YSC]

Someone was going to loan me one but backed out due to shipping costs. I can buy one but I have bought so many already that I need to test first.

Too bad for that as there in house measurement was incredible for such a cheap speaker I actually am tempted to upgrade my kef x300a to the t5v as my desktop speaker in tiny room here in Hong kong

 

[pma]

We also have to remember that this is an MSRP $900 speaker. No one should pay anywhere near that for them. $6-$700 on a bad day is more realistic. .

In my country, a price of the pair starts at 29 900,- CZK, which is the equivalent of $1329 at the moment. This is quite a lot of money for a bookshelf small speakers.

 

[ROOSKIE]

You're talking like these come with a free sub; and the money to move if you live in a not so well isolated appartment. Sure, these are supposed to be used with a subwoofer, but there are cases where they'll be used without.
I mean, they would have used a 4" or 5" woofer if a subwoofer was mandatory; makes for a better midrange and cheaper speaker.

I think these have plenty of bass for their size and reasonable sensitivity.
If you prefer lower extension there are other options but likely tradeoffs in other areas.
6" is good size for dynamic response and likely allows for a lower crossover point. I prefer crossing lower than 80hrz/100hrz which is the limit for bass localization.

 

[napilopez]

The dips in the vertical response around the crossover point have nothing to do with floor and ceiling reflections. These are indicators of the lobe behavior - a wrong observation that has been made in several of the other speaker reviews. The woofer is sourced from China, not Indonesia (SB Acoustics).

Hi @Selah Audio. The floor and ceiling bounce curves are part of the spinorama/CTA-2034A standard, components of the early reflections curve. Amir isn't saying these dips are caused by the reflections, but rather that this is the sound that will hit your floor and ceiling and bounce back at you in a typical home. The dips are of course precisely as you say, because of the lobe behavior.

In case anyone else is confused by this, the floor bounce is an average of the measurements at -20, -30, and -40 degrees vertical, and the ceiling bounce is an average of +40,+50, and +60 curves. These angles were determined from the 2002 devantier paper as being perceptually important in the average home. Of course, the reality will depend on your individual home and speaker placement, but they should be representative for a wide swath of people.

 

[Soundstage]

In my country, a price of the pair starts at 29 900,- CZK, which is the equivalent of $1329 at the moment. This is quite a lot of money for a bookshelf small speakers.

Kef LS50 are now sold for USD 650. Choose your weapon!

 

[QMuse]

Where are you getting the idea they are not? Nearly every Spin I've seen of a decent speaker has the two lines extremely close together. There are exceptions (tweeter/waveguide bump/cancellation that happens only exactly on axis), but those are the exceptions. "Generally," the lines are nearly on top of one another and if they're not it's probably not a very good speaker.

From the way they are measured and calculated. On-axis is measured with single sweep made at 0° while Listening Window is an average of that same on-axis sweep and 8 more measuremens (±10°,±20° & ±30° in the horizontal plane and ±10° in the vertical plane). As such Listening Window is much more relevant than a single on-axis sweep as it contains much more spatial information. Sure, with most quality speakers they are close to each other as quality speakers have smooth off-axis response near the on-axis, but that still cannot be used as an argument that those 2 curves are the same as Listening Window provides much more information about speakers (near) on-axis response than a single on-axis sweep.

 

[Putter]

Sure, but this was F206/F208. They have a more advanced cone with "Geometrically reinforced aluminum cone for optimum piston operation and reduced break-up". You can easily see the reinforced points on the cone. They also use a 5" for midrange, I believe it breaks up around 8khz.

The M16 on the other hand uses a simpler cone and likely has a much lower breakup point. I also find it incredibly unlikely that it has a 4th order electrical crossover given the price of the speaker. It might be something resembling 4th order acoustically, but in that case the breakup would be noticeable (but not necessarily very audible) without a notch filter.

I find it incredibly likely that it does have a 4th order crossover. From a Stereophile review of its poor relation, the Infinity Primus 150 Crossover slopes: 24dB/octave. Price: $198/pair. The OM for the Primus line indicates 4th orders slopes for the whole series.

This is a picture of the Infinity Primus 163. I'm not an expert on crossovers, but it's definitely not a 1st order. It does use electrolytic caps and iron core inductors as cost saving measures which I suspect isn't the case for a luxury brand.

 

[QMuse]

LW is the average of 9 curves: on-axis, ±10, 20 and 30° horizontal, and ±10° vertical. So on-axis contributes 12/9 ~1.3% to PIR. This is deceptive though, as these curves are mostly quite collinear variables (high correlation with each other), with the small-angle curves being particularly similar to the on-axis response in most cases, so the 'effective' contribution of the on-axis response to LW (and so PIR), directly plus via its correlation with the off-axis curves, is higher than it would be if the curves were true independent variables.

All 9 curves are true independent variables and are specific to the speaker been tested. There is no border angle from which curves magically become independent but are dependent with lower angles. Sure, with quality speakers all 9 curves used to calculate LW would be similar to on-axis, but that doesn't mean that LW is the same thing as on-axis curve as LW does contain 8 additional spatial information which on-axis curve doesn't have. this being much mroe valuable.

So even if both curves measure exactly the same they still don't provide the same information.

 

[March Audio]

Compared to what? I was under the impression that anything under 5-10% in the LF region is undetectable?

https://books.google.com/books?id=2m9DNXs05mUC&pg=PA12&lpg=PA12&dq=percent+of+distortion+that+is+detectable+in+speakers&source=bl&ots=ZEmFc3ofNY&sig=ACfU3U37WNhuAuPdjx4bst9Qa4zX5-Vkrg&hl=en&sa=X&ved=2ahUKEwjJiMTExoToAhXOhHIEHfRvDYkQ6AEwCnoECAkQAQ#v=onepage&q=percent of distortion that is detectable in speakers&f=false

You are kidding right?

You didn't notice the spike at 300hz and the big issue as it approaches the crossover point at 2.1kHz. If that is really cone breakup I am really surprised Harman thought it was acceptable.

Compared to? Many of the other speakers tested.

 

[maty]

...This is a picture of the Infinity Primus 163. I'm not an expert on crossovers, but it's definitely not a 1st order. It does use electrolytic caps and iron core inductors as cost saving measures which I suspect isn't the case for a luxury brand.

View attachment 53028

Revel Performa3 series are more expensive but...
https://www.audiosciencereview.com/...idence-based-speaker-designs.6441/post-153068

 

[maty]

https://www.revelspeakers.com/info/series-performa3.html

https://6moons.com/audioreviews/revel2/2.html

and cement / sand resistors too.

That said, the audio industry only spend on the external appearance.

 

[YSDR]

With LR4 at 2100Hz there is 37dB of attenuation at 6kHz.

LR4 slopes have visible contribution to the summed frequency response at more than 1 octave from the crossover point if phase tracking is good. So the peak around 5 kHz it's very likely from the woofer cone break-up, the 3rd harmonic distortion rise between 1-2 kHz confirms this (5000/3=1666 as the start of the rise). The resonance is visible at the waterfall plot too.
Often even luxury manufacturers save on their cheaper models, which I don't think is surprising to anyone. A notch filter was omitted from this speaker's crossover or the break-up point of the woofer was moved during production spraying or so.

 

[QMuse]

LR4 slopes have visible contribution to the summed frequency response at more than 1 octave from the crossover point if phase tracking is good. So the peak around 5 kHz it's very likely from the woofer cone break-up, the 3rd harmonic distortion rise between 1-2 kHz confirms this (5000/3=1666 as the start of the rise). The resonance is visible at the waterfall plot too.

Sure, those incompetent guys from harman simply screwed it. I hope they will learn from your post and make a better speakers in the future.

 

[YSDR]

Sure, those incompetent guys from harman simply screwed it. I hope they will learn from your post and make a better speakers in the future.

Sorry, your comment is laughable. Of course, Harman have great knowledge about speaker building, but they don't necessarily want to pack all their knowledge to the cheaper models, this is economy.
So overall, you defend your statement with that "Harman knows everything and does it at all price level because they really so good guys to us".
I am wondering how many drivers you measured or how many speakers you built.

 

[QMuse]

Sorry, your comment is laughable. Of course, Harman have great knowledge about speaker building, but they don't necessarily want to pack all their knowledge to the cheaper models, this is economy.
So overall, you defend your statement with that "Harman knows everything and does it at all price level because they really so good guys to us".
I am wondering how many drivers you measured or how many speakers you built.

Let me draw it for you. On the right side of the yellow vertical line (which is at XO point at 2100Hz) you can see a curve which shows contribution of the woofer after the XO point. On the left side of the yellow curve same thing for tweeter is shown. As you can see it is at -30dB at 5000Hz so cone breakup would be really well attenuated. Which can also be seen from predicted in-room response as there is no resonance peak at 5000Hz. In fact, there is no resonance peak anywhere on that curve.

But sure, you would know better as you measured like 300 drives and built like 50 different types of speakers, right?

 

[maty]

Would not it be better to ask Harman or Revel directly?

https://www.revelspeakers.com/contact.html

 

[YSDR]

It would be great if the speaker would have bi-wire option, then we can see clearly that this is a cone breakup or not. The measurements tells, it's an non-attenuated (other than crossover slope) cone breakup. Even a 10-15 dB breakup resonance peak is possible. And as I said, even a breakup move is possible during manufacturing. So let's say the prototype breakup was at 6 khz, Harman put a 6 Khz notch to the crossover network and then for some reason the breakup moved to 5 kHz due to manufacturing tolerances or so, and suddenly the notch is not beneficial but even more problematic as if it wasn't there.

 

[YSDR]

Would not it be better to ask Harman or Revel directly?

https://www.revelspeakers.com/contact.html

You can't be so naive. If something goes wrong with the speaker or simply they saved production costs, would they tell us?
Of course, we have to assume that they tell us ...